Documentation

Guides, API reference, and integration setup for the MerchantGuard™ platform. Connect via MCP, REST, or our SDKs. Built on 4 filed U.S. provisional patents and an open-source SDK family.

Get API Key →

Getting Started

MerchantGuard ships compliance infrastructure for high-risk merchants and AI-agent commerce. Connect via remote MCP server, REST endpoints, or our published SDKs. Get your API key at /developers and pick an integration below.

Claude Desktop

Settings → Connectors, or config file + Node

Cursor

Tools & MCP or mcp.json — all platforms

Claude.ai (Web)

Settings → Connectors, paste remote MCP URL

OpenClaw

Plugin with auto-recall and capture

Claude Code

One-line claude mcp add command

Local LLM

Ollama, LM Studio, Qwen — fully air-gappable

REST API

Direct HTTP endpoints

Python SDK

pip install merchantguard-sentinela

JavaScript SDK

npm install @merchantguard/guard

Featured products

The shipped MerchantGuard family. Open-source where it makes sense, sales-led where compliance demands it.

AgentGuard® Chargeback

Open-source CE 3.0 evidence SDK

GuardScore™

FICO-style compliance score 0-100

Sentinela

Conversational intent firewall, <10ms p95

GuardGate

Agent authorization firewall

Mystery Shopper

10-probe adversarial agent audit

GuardScan

3-layer security scanner

FleetScore

Agent fleet monitoring

AI Compliance Coaches

14 vertical playbooks

Trust posture

MerchantGuard is the compliance layer for high-risk merchants and AI-agent commerce. USPTO trademarks pending: MerchantGuard™ (Serial 99051215), GuardScore™ (Serial 99030125), AgentGuard® (Serial 99462472). Four U.S. provisional patent applications filed Feb 17 2026: 63/983,615 / 63/983,621 / 63/983,843 / 63/984,626. Defensive patent pledge modeled on LOT Network and Open Invention Network. Read LEGAL.md, PATENTS.md, and DISCLAIMER.md for the full posture.